Proc. Natl. Acad. Sci. USAVol. 93, pp. 11448-11453, October 1996Applied Biological Sciences

Overexpression of lecithin:cholesterol acyltransferase intransgenic rabbits prevents diet-induced atherosclerosis

(high density lipoprotein cholesterol/low density lipoprotein cholesterol/atherogenic diet/cholesteryl esters)

JEFFREY M. HOEG*t, SILVIA SANTAMARINA-FOJO*, ANNIE M. BE'RARD*, J. FREDRICK CORNHILL*,EDWARD E. HERDERICKt, SANFORD H. FELDMAN*, CHRISTIAN C. HAUDENSCHILD§, BORIS L. VAISMAN*,ROBERT F. HOYT, JR.*, STEPHEN J. DEMOSKY, JR.*, RICHARD D. KAUFFMAN*, CHRISTINA M. HAZEL*,SANTICA M. MARCOVINAS, AND H. BRYAN BREWER, JR.**Molecular Disease Branch and the Laboratory of Animal Medicine, Building 10, Room 7N115, 10 Center Drive, MSC 1666, National Heart, Lung, and BloodInstitute, National Institutes of Health, Bethesda, MD 20892-1666; tBiomedical Engineering Center, 270 Bevis Hall, 1080 Carmack Road, Ohio State University,Columbus, OH 43210; §The Holland Laboratory, The American Red Cross, 15601 Crabbs Branch Way, Rockville, MD 20855; and INorthwest Lipid ResearchLaboratories, University of Washington, Seattle, WA 98103

Communicated by William H. Daughaday, Balboa Island, CA, May 7, 1996 (received for review March 3, 1996)

ABSTRACT Lecithin:cholesterol acyltransferase (LCAT)is a key plasma enzyme in cholesterol and high densitylipoprotein (HDL) metabolism. Transgenic rabbits overex-pressing human LCAT had 15-fold greater plasma LCATactivity than nontransgenic control rabbits. This degree ofoverexpression was associated with a 6.7-fold increase in theplasma HDL cholesterol concentration in LCAT transgenicrabbits. On a 0.3% cholesterol diet, the HDL cholesterolconcentrations increased from 24 ± 1 to 39 + 3 mg/dl innontransgenic control rabbits (n = 10; P < 0.05) and in-creased from 161 + 5 to 200 ± 21 mg/dl (P < 0.001) in theLCAT transgenic rabbits (n = 9). Although the baselinenon-HDL concentrations of control (4 ± 3 mg/dl) and trans-genic rabbits (18 ± 4 mg/dl) were similar, the cholesterol-richdiet raised the non-HDL cholesterol concentrations, reflect-ing the atherogenic very low density, intermediate density, andlow density lipoprotein particles observed by gel filtrationchromatography. The non-HDL cholesterol rose to 509 + 57mg/dl in controls compared with only 196 ± 14 mg/dl in theLCAT transgenic rabbits (P < 0.005). The differences in theplasma lipoprotein response to a cholesterol-rich diet observedin the transgenic rabbits paralleled the susceptibility to devel-oping aortic atherosclerosis. Compared with nontransgenic con-trols, LCAT transgenic rabbits were protected from diet-inducedatherosclerosis with significant reductions determined by bothquantitative planimetry (-86%; P < 0.003) and quantitativeimmunohistochemistry (-93%;P < 0.009). Our results establishthe importance of LCAT in the metabolism of both HDL andapolipoprotein B-containing lipoprotein particles with choles-terol feeding and the response to diet-induced atherosclerosis. Inaddition, these findings identify LCAT as a new target fortherapy to prevent atherosclerosis.

The development of human atherosclerosis is inversely relatedto the concentration of the high density lipoproteins (HDL)(1). High and low concentrations of plasma HDL are associ-ated with decreased and increased risk of developing prema-ture atherosclerotic cardiovascular disease, respectively (2). Ithas been proposed that a 1% increase in the concentration ofHDL would lead to a 3% reduction in risk for developingclinical atherosclerosis in man (1). An important enzyme inmodulating plasma HDL concentrations is lecithin:cholesterolacyltransferase (LCAT). LCAT catalyzes the transfer of fattyacid from the sn-2 position of lecithin to the free hydroxylgroup of cholesterol and was first proposed nearly 30 years agothat cholesterol esterification would be a key step in transfer-

ring cholesterol from the tissues of the body to the liver (3).This process, termed reverse cholesterol transport (4), wasproposed to facilitate the removal of cholesterol from thebody. Thus reverse cholesterol transport is one of severalproposed mechanisms by which HDL provides protection fromcardiovascular disease (5-10).

Inborn errors of metabolism can point to the in vivo effectsof either overexpression or loss of enzymatic function. Recentstudies have established mutations in the gene encoding LCATthat lead to the total or partial loss of LCAT activity in theplasma (11-14). This leads to reduced concentrations of HDLcholesterol and accumulation of cholesterol in the kidney andcornea (14). To assess the impact that overexpression ofLCAThas on the plasma lipoproteins, we and others have generatedtransgenic mice (15, 16) and rabbits (17) expressing high levelsof human LCAT. In both species, human LCAT raises theplasma HDL cholesterol concentrations. In addition, the con-centration of the non-HDL lipoprotein particles was unexpect-edly reduced in transgenic rabbits expressing high levels ofhuman LCAT (17). Because the cholesterol-fed rabbit is theclassic model for the study of diet-induced atherosclerosis (18),we have tested the hypothesis that overexpression of LCAT incholesterol-fed rabbits would not only increase the HDLcholesterol concentrations in these animals but also preventthe development of atherosclerosis. In addition to raisingplasma HDL cholesterol concentrations, the concentrations ofthe atherogenic very low density lipoprotein (VLDL), inter-mediate density lipoprotein (IDL), and low density lipoprotein(LDL) particles were substantially lower in the cholesterol-fedLCAT transgenic rabbits than the concentrations in the con-trols. These results indicate that LCAT may protect againstdiet-induced atherosclerosis by modulating both HDL andnon-HDL particle metabolism.

MATERIALS AND METHODSControl and Transgenic Animals. The full-length human

LCAT gene was used to generate transgenic animals (15, 17).This genomic construct contained all of the introns and 851 bpof the 5' and 1134 bp of the 3' untranscribed regions of thehuman LCAT gene. The generation of transgenic rabbits wasapproved by the Animal Care and Use Committee of theNational Heart, Lung, and Blood Institute. New ZealandWhite rabbits were purchased from Hazelton Research Prod-ucts (Denver, PA). Rabbits were housed in separate cages. A

Abbreviations: LCAT, lecithin:cholesterol acyltransferase; HDL, highdensity lipoprotein; IDL, intermediate density lipoprotein; LDL, lowdensity lipoprotein; VLDL, very low density lipoprotein; CETP,cholesteryl ester transfer protein.tTo whom reprint requests should be addressed.

11448

The publication costs of this article were defrayed in part by page chargepayment. This article must therefore be hereby marked "advertisement" inaccordance with 18 U.S.C. §1734 solely to indicate this fact.

Proc. Natl. Acad. Sci. USA 93 (1996) 11449

total of 9 LCAT transgenic and 10 control males 5-6 monthsof age were studied.Plasma LCAT, Lipid, and Lipoprotein Analyses. Blood was

drawn for analysis in tubes containing EDTA from rabbitsafter a 12-h fast. Plasma a-LCAT activity was determinedusing 10 ,ul of plasma in a proteoliposome assay (19). CETPactivity was determined as described by Albers, Chen, andcoworkers (20), and post-heparin hepatic lipase activity wasdetermined as described by Iverius and Brunzell (21). The totalcholesterol and triglyceride concentrations (Sigma) as well asunesterified cholesterol and phospholipids (Wako Chemicals,Richmond, VA) were determined on plasma samples usingenzymic methods on a Hitachi 911 Autoanalyzer (BoehringerMannheim). The esterified cholesterol was calculated by sub-tracting unesterified from the total cholesterol. The non-HDLcholesterol concentration was determined by quantitating thecholesterol concentration in the supernatant of plasma that hadbeen diluted with PBS (1:1 vol/vol) and then precipitated withdextran sulfate (22). The HDL cholesterol was calculated bysubtracting the cholesterol concentration in the supernatant afterprecipitation from the total plasma cholesterol concentration.Gel filtration chromatography was performed on plasma

samples as described (17). Briefly, 200 ,lJ of plasma wasapplied to a fast protein liquid chromatography system con-sisting of two Superose 6 columns in series (Pharmacia) andeluted with 1 mM EDTA and 0.02% (wt/wt) sodium azide inPBS (23). The first 10 ml of eluate was discarded, and theremaining 30 ml, containing the plasma lipoprotein fractions,was collected in 0.5-ml aliquots. The total cholesterol, unes-terified cholesterol, phospholipid, and triglyceride concentra-tions of each of these 0.5-ml fractions were determined asoutlined above.

Diet-Atherosclerosis Study. Rabbits were fed a daily rationof 120 g of 0.3% cholesterol diet (Ziegler Brothers, Gardener,PA; product number 4109000). After 17 weeks, the rabbitswere killed using intravenous sodium pentobarbital. The aor-tas were harvested and stained with sudan IV, and thepercentage of the surface area stained was determined byplanimetry of the digitized image (24). One-millimeter slices ofaorta just inferior to the left subclavian artery ostia wereharvested from each aorta, stained, and analyzed. The extentof intimal cellular proliferation was quantitated using a ratioof the intima to media (25).

Hepatic Lipid and Cholesterol Quantitation. After harvest-ing of the aorta, approximately 200 mg of liver (wet weight) wasweighed, minced, and extracted in 20 ml of chloroform/methanol (2:1, vol/vol) using the method by Folch and co-workers (26). After drying of the organic phase under nitro-gen, lipid content was determined gravimetrically, and thelipids were resolubilized in 2 ml of isopropanol. Proteincontent was quantitated on the remaining tissue after over-night solubilization in 5 ml of NaOH (27) using the enhancedprotocol of the bicinchoninic acid method (Pierce). Totalcholesterol content was assayed using the Cholesterol CIIEnzymatic Colorimetric Method (Wako Chemicals). Unester-ified cholesterol content was ascertained using the Free Cho-

lesterol C Enzymatic Colorimetric Method (Wako Chemicals).Esterified cholesterol was calculated by subtraction.

RESULTSPlasma LCAT activity was 101 ± 11 nmol/ml/h in thenontransgenic control rabbits. In contrast, the LCAT activityin the LCAT transgenic rabbits was 1593 ± 101 nmol/ml/h andwas significantly higher than controls (P < 0.0001). Comparedwith controls, LCAT transgenic rabbits had a marked increasein both total (617%; P < 0.001) and HDL cholesterol con-centrations (671%; P < 0.001; Table 1). On the cholesteroldiet, control rabbits had 19-fold and 127-fold increases in totaland non-HDL cholesterol concentrations, respectively, as wellas a 74% increase in the plasma triglyceride concentrations(Table 1). In contrast, the plasma total and non-HDL choles-terol concentrations in the LCAT transgenic rabbits increasedonly 2-fold and 11-fold, respectively, on the cholesterol diet.The LCAT activity in the transgenic rabbits on the cholesteroldiet remained more than 3-fold that of controls and wasassociated with an increase of HDL cholesterol concentrationto more than 5-fold that of control rabbits. Analysis of theplasma lipoproteins by gel filtration chromatography demon-strated that cholesterol-fed control rabbits had cholesterolprincipally in VLDL, IDL, and LDL particles (Fig. 1). Incontrast, a substantial fraction of the plasma cholesterol in theLCAT transgenic rabbits was present in large HDL particles(Fig. 1A). The majority of the cholesterol in both controls andLCAT transgenic rabbits was esterified (Fig. 1B). However,compared with nontransgenic control rabbits, the HDL in theLCAT transgenic rabbits were enriched in both triglyceridesand especially phospholipids (Fig. 1 C and D). These differ-ences in the plasma lipoproteins were reflected in the total/HDL cholesterol ratio. The total cholesterol/HDL cholesterolratio, a sensitive indicator of clinically detectable humanatherosclerosis (28), increased in the control group by morethan 12-fold. In contrast, in LCAT transgenic rabbits, thetotal/HDL ratio rose less than 2-fold (Table 1) and remainedbelow the ratio of 5, which provides an average risk foratherosclerosis in man (29).Both the CETP and post-heparin hepatic lipase activities

were determined in these rabbits. On a regular chow diet, theCETP activity (expressed as percent per 5 ,ul per 18 h) in theLCAT transgenic rabbits (30 ± 2) was twice that of controls(15 ± 1; P < 0.001). In addition, there was a significantdifference in hepatic lipase activity. The activity of post-heparin hepatic lipase in controls (56 ± 4 nmol/ml/h) wassignificantly higher than in the LCAT transgenics (34 ± 1nmol/ml/h; P < 0.02). Cholesterol feeding increased theactivities of both of these plasma proteins in both control andtransgenic rabbits. The CETP activity in control rabbits in-creased to 35 + 3, and the LCAT transgenic rabbit CETPactivity increased to 58 ± 4. These values from cholesterol-fedanimals were significantly higher than those from chow-fedanimals in both strains of rabbit (P < 0.001), and the LCATtransgenic CETP activity remained 66% higher than controls(P < 0.005). The hepatic lipase activities were also significantly

Table 1. Plasma lipoproteins before and after cholesterol feeding in control and LCAT transgenic rabbits

Total Total HDL non-HDL Totalcholesterol, triglycerides, cholesterol, cholesterol, cholesterol/HDL

mg/dl mg/dl mg/dl mg/dl cholesterol, mg/dlControl (n = 10)

Baseline 29 ± 3 39 ± 4 24 ± 1 4 ± 3 1.17 ± 0.12Cholesterol-fed 548 ± 57* 107 ± 15* 39 ± 3* 509 ± 57* 14.98 ± 2.13*

LCAT-transgenic (n = 9)Baseline 179 ± 7t 43 ± 4 161 ± St 18 ± 4 1.11 ± 0.02Cholesterol-fed 396 ± 33*t 81 ± 8*t 200 ± 21*t 196 ± 14*t 2.03 ± 0.07*t

*Differs from baseline; P < 0.05.tDiffers from control values; P < 0.05.

Applied Biological Sciences: Hoeg et al.

11450 Applied Biological Sciences: Hoeg et al.

Elution Volume (ml)

FIG. 1. Gel filtration chromatography of plasma from control(solid lines) and transgenic (dashed lines) rabbits ingesting a 0.3%cholesterol-enriched diet. The concentrations of cholesterol (A),cholesteryl ester (B), triglycerides (C), and phospholipids (D) weredetermined on 0.5-ml fractions eluted from the two sequential Sepha-rose 6 columns.

higher in both control (95 ± 7) and LCAT transgenic animals(61 ± 5) after cholesterol feeding (P < 0.005). However, thehepatic lipase remained significantly higher in the controls thanin the transgenic rabbits (P < 0.001). Therefore, LCAT overex-

pression was associated with substantial changes in plasma activ-ity of both of these proteins that affect HDL metabolism.To determine if cholesterol feeding was associated with

changes in hepatic cholesterol content, the livers of control andLCAT transgenic rabbits were analyzed. LCAT overexpressiondid not alter the hepatic lipid content compared with controlrabbits (Table 2). Whether expressed per mg of tissue or permg of hepatic protein content, the total lipid, cholesterol, andcholesteryl ester contents in rabbit liver were not differentbetween control and LCAT transgenic rabbits (Table 2). Inaddition, the histology of the livers of control and LCATtransgenic rabbits were similar and disclosed only mild hepa-tocytic lipid accumulation. Finally, serum transaminase con-centrations were not affected with cholesterol feeding in eithercontrol or LCAT transgenic rabbits (data not shown). There-fore, the changes observed in the plasma lipoproteins withLCAT overexpression were not secondary to hepatotoxicityand did not lead to alterations in hepatic cholesterol content.To evaluate the potential protective role of LCAT overex-

pression in the development of diet-induced atherosclerosis,control and LCAT transgenic rabbits were placed on a 0.3%cholesterol diet. After 17 weeks on the 0.3% cholesterol diet,the aortas from control and LCAT transgenic rabbits wereharvested, and two methods were used to quantitate theseverity of diet-induced atherosclerosis. Sudan IV staining ofthe lipid droplets was used to quantify the percent of thesurface area developing lesions (24). The probability map foraortic lesion development in the transgenic rabbits showedonly scattered foci of sudan IV-staining material, whereascontrol aortas had substantial staining in the majority of theanimals (Fig. 2). The aortas of the control group had 35 ± 7%of the surface covered by plaque. In marked contrast, only 5 ±1% of the aortic surface was covered by plaque in the LCATtransgenic rabbits (P < 0.009; see Figs. 2 and 4).The substantial differences in the atherosclerosis in aortas

between control and LCAT transgenic rabbits were alsoevident microscopically (Fig. 3). The intima of the controlrabbits demonstrated foam cell formation, cellular prolifera-tion, and an increase in the ratio of the intima/media to 0.40 ±0.11 (Fig. 3 Left). There was virtually no foam cell formationor cellular proliferation in the transgenic rabbits expressinghuman LCAT (see Figs. 3 and 4 Right). The 0.03 ± 0.01intima/media ratio was significantly lower than the control(P < 0.009; Fig. 4). Therefore, LCAT overexpression led to an85-90% reduction in diet-induced atherosclerosis in rabbits.The atherosclerosis assessed by two different end points

used in this study were highly correlated for both the controlrabbits (Fig. 5A; r = 0.87; P < 0.001) as well as for the entirestudy group (r = 0.92; P < 0.001). The severity of atheroscle-rosis was inversely correlated with the plasma LCAT activity(r = -0.55; P < 0.019). This was true not only with LCAToverexpression but also within the control group. The higherthe LCAT activity in the nontransgenic control rabbits, thelower the extent of atherosclerosis (r = -0.64; P < 0.006; Fig.SB). With LCAT overexpression, both the non-HDL choles-terol (r = 0.82; P < 0.0001) and the total cholesterol/HDLcholesterol ratio (r = 0.89; P < 0.001) correlated with theseverity of atherosclerosis. However, the intima/media ratioand the total/HDL cholesterol concentration were also highlycorrelated within the nontransgenic control group (Fig. 5 C

Table 2. Hepatic lipid content in control and LCAT transgenic rabbits after cholesterol feeding

Per mg of tissue (wet weight) Per mg of cell protein

Total Cholesteryl Total CholesterylTotal lipid, cholesterol, ester, Total lipid, cholesterol, ester,mg/mg pxg/mg pg/mg mg/mg tg/mg ,g/mg

Control (n = 10) 0.12 ± 0.03 28.5 ± 7.0 20.8 ± 5.1 0.87 ± 0.28 206.4 ± 63.0 149.9 + 45.2Transgenic (n = 9) 0.09 ± 0.03 22.4 ± 11.8 16.8 ± 9.5 0.64 + 0.29 161.4 ± 93.7 121.7 + 74.7

Proc. Natl. Acad. Sci. USA 93 (1996)

Proc. Natl. Acad. Sci. USA 93 (1996) 11451

Control (n=1 0)35% ± 7%

0-10 10-20 20-30 30-40 40-50 50+

FIG. 2. Comparison of the atherosclerosis in control and transgenic rabbits overexpressing human LCAT by quantitative planimetry. The aortasof LCAT transgenic (n = 9) and control (n = 10) male rabbits, fed a 0.3% cholesterol chow diet for 17 weeks, were harvested and stained withsudan IV. The compilations of the images from the study groups are summarized for transgenic and control rabbits with a graded coloration ofthe probability of distribution shown at the bottom.

and D). These dose-response relationships in both LCAToverexpressors and the nontransgenic control group furtherstrengthen the association between diet-induced atherogenesisand the level of LCAT expression.

DISCUSSIONThis study was undertaken to evaluate the role of LCAToverexpression in modulating the plasma lipoproteins andatherosclerosis. Transgenic rabbits overexpressing humanLCAT were generated, and the development of atherosclerosisin rabbits was evaluated after the animals were fed a highcholesterol diet. Rabbit lipoprotein metabolism is unique forseveral reasons. These animals have elevated concentrations ofplasma CETP (30), they do not synthesize apolipoprotein A-II(31-33), and they have reduced plasma hepatic lipase activity(34) compared with murine species. The feeding of cholesterolto rabbits has long been known to lead to atherosclerosis (18).High cholesterol diets lead to the accumulation of lipid inapolipoprotein B-containing VLDL, IDL, and LDL particlesin control rabbits (Fig. 1). The increase in plasma concentra-tions of non-HDL lipoprotein particles in control rabbits wasconsiderably greater than in the transgenic animals overex-pressing LCAT after cholesterol feeding (Fig. 1 and Table 1).Non-HDL cholesterol concentrations increased in LCAT

transgenic rabbits to only 39% of that observed in nontrans-genic controls. In contrast, the plasma HDL concentrationswere 5-fold higher in the LCAT-expressing strain than incontrol rabbits. These differences in the plasma lipoproteinswere associated with resistance to diet-induced atherosclerosisin the LCAT transgenic rabbits. Compared with nontransgeniccontrols, the LCAT-expressing strain experienced an 86% reduc-tion in the surface of the aorta covered by plaque (Figs. 2 and 4)and 93% reduction in the intima/media ratio (Figs. 3 and 4).LCAT overexpression led to marked changes in both the plasmalipoproteins and in cholesterol-induced atherogenesis.LCAT overexpression had a variety of effects on the impact

of a cholesterol-rich diet on the plasma lipoproteins, leading toan antiatherogenic lipoprotein profile. The elevated concen-trations of HDL observed in chow-fed LCAT transgenicrabbits (17) were even more pronounced with cholesterolfeeding (Table 1 and Fig. 1). The HDL cholesterol concen-trations with cholesterol feeding in LCAT transgenic rabbits(200 ± 39 mg/dl) was 5.1-fold that in nontransgenic controls(39 ± 3 mg/dl). The large HDL particles generated in theLCAT transgenic rabbit contained not only cholesteryl ester(Fig. 1B) generated from the LCAT reaction, but theselipoproteins also had a high triglyceride content (Fig. 1C). Thistriglyceride enrichment reflected, at least in part, the 66%

Control TransgenicFIG. 3. A 1-mm section was taken at the descending thoracic aorta at the same position for each aorta and stained with periodic acid/Schiff

reagent, and the degree of foam cell accumulation in the intima in the controls (Left) was compared with the lack of intimal foam cell formationor change in intimal thickness in the transgenic aortas (Right).

Applied Biological Sciences: Hoeg et al.

11452 Applied Biological Sciences: Hoeg et al.

Controls Transgenics

FIG. 4. The extent of intimal cellular proliferation as shown in Fig.2 was quantitated using a ratio of the intima to media (25). Thequantitative assessment of both the intima/media ratio (P < 0.003)and the percent of the surface area (P < 0.009) was significantly lowerin the transgenic LCAT rabbits than in controls.

higher CETP activity and the 31% reduction in hepatic lipaseactivity in the LCAT transgenic rabbits compared with con-trols. CETP exchanges cholesteryl ester from the core of HDLparticles with triglyceride from the core of apolipoproteinB-containing particles (35, 36). The large HDL particle size,triglyceride content, and particularly the high phospholipidconcentration in HDL (Fig. 1D) reflect the lower activity ofhepatic lipase in LCAT transgenic rabbits.

I/M

1.00

I/M

AB0 _ r= 0.87 * .

p < 0.001

30~~~~~~~

10 0

0-I I~~~~~~~~~~~

0 10 20 30 40 50 60 70 81% Surface Area

The elevated plasma levels ofHDL and reduced apolipopro-tein B-containing lipoproteins in the LCAT transgenic rabbitswere associated with a markedly decreased lipid accumulationand cellular proliferation in the aortic intima (Figs. 2-4). Bothmethods for quantitation of atherogenesis employed in thesestudies, the extent of the aortic surface covered by plaque andthe intima/media ratio, were significantly reduced in theLCAT transgenic rabbits. The decreased aortic cellular lipidand cholesterol accumulation may be due to both the reductionin the proatherogenic apolipoprotein B lipoproteins as well asthe increased HDL lipoproteins. Several lines of evidencesupport the role that HDL plays in modulating the develop-ment of atherosclerosis. Direct infusion of HDL intravenouslyhas been shown to attenuate atherogenesis in lagomorphs (37).Raising HDL concentrations in apolipoprotein A-I transgenicmice resulted in decreased diet-induced atherosclerosis (38).In addition, transgenic apolipoprotein A-I mice crossed withapolipoprotein E knockout mice led to mice with raised plasmaHDL levels and protection from the spontaneous atheroscle-rosis that develops in apolipoprotein E-deficient mice (39).These observations in animal models may also apply to man,since overproduction of apolipoprotein A-I has also beenreported to be associated with hyperalphalipoproteinemia andenhanced longevity (40).A number of mechanisms by which HDL could mediate

protection from atherosclerosis have been proposed. HDLhave been proposed to both induce cholesterol efflux (41, 42)and serve as a mitogen for cellular proliferation (43). HDL canalso serve as a buffer to protect apolipoprotein B particlesfrom potentially proatherogenic oxidation (44). HDL is het-

I/M

1.01.9'.81.71.61.51

.31

.21

0

100 200 300 400 500 600 700 800 900

NonHDL(mg/dl)

o B~oO 01000,oO -

r = -0.64 0 \o p C 0.0450O ' ' 1 1 170 80 90 100 110 12

0

LCAT Activity(nmol/ml/h)

!0

5 10 15 20 25 30

TC/HDL

FIG. 5. Significant (P < 0.05) bivariate Pearson correlations of post-diet variables with aortic atherosclerosis in nontransgenic, control rabbits.The intima/media (I/M) ratio correlated well with the planimetry assessment of atherosclerosis (A). The I/M ratio was inversely correlated withLCAT activity (B) and positively correlated with the non-HDL cholesterol (C) and total cholesterol/HDL cholesterol (TC/HDL) (D).

0I=

cu

C._3

cUn

a)

G)

cnU

10

CD)

Controls Transgenics

D.80 r= 0.86 *

p C 0.001.60 -

I/M.40 _

.20 -

0 _ /

_In I

Proc. Natl. Acad. Sci. USA 93 (1996)

*'

*:

Proc. Natl. Acad. Sci. USA 93 (1996) 11453

erogeneous and contains several different lipoprotein parti-cles. In man, the HDL-containing apoA-I (LpA-I) has beenproposed to be more effective in reverse cholesterol transportthan particles containing both apoA-I and apoA-II (45-47).ApoA-I is a potent cofactor enhancing LCAT activity, and themodulation of LpA-I size is sensitive to the presence ofapoB-containing particles and LCAT activity (48). Becauserabbits express no apoA-II (31-33), these transgenic rabbitshave only LpA-I particles. Alternatively, the reduced concen-tration of atherogenic apolipoprotein B particles with LCAToverexpression may reduce the endothelial damage and cel-lular proliferation induced by IDL and LDL (49).The effects of LCAT overexpression on atherogenesis was

significantly correlated with the changes in the plasma lipopro-teins. However, these study end points, which were so strikingin the LCAT transgenic rabbits, were also apparent within thenontransgenic control rabbits. The LCAT activity in thecontrol group correlated with the total cholesterol/HDLcholesterol ratio (Fig. SD) and the severity of atherosclerosis(Fig. 5A and C). Therefore, the changes observed in the LCATtransgenics may reflect a more subtle impact (i.e., LCAT mayhave on the plasma lipoproteins and susceptibility to athero-genesis at more physiologic levels of expression).The findings in this study may be of relevance to athero-

genesis in man. Although a deficiency in the plasma LCATactivity was not originally associated with atherosclerosis (50),some mutations in the LCAT gene may be more deleterious toapolipoprotein B metabolism and lead to enhanced cardio-vascular disease risk (51). However, the therapeutic impact ofgene overexpression may not be predicted by the clinicalobservations in inborn errors in metabolism. Modulation oflipoprotein metabolism relevant to the prevention of athero-genesis by LCAT may occur in the context of other regulatorysteps within the circulation. This may be the case in the generalpopulation. Adults aged 20-59 years in the Pacific NorthwestBell Telephone Company study were found to have plasmaLCAT masses ranging from 2.87 to 8.56 ,ug/ml (52). This studyindicated that women and nonsmokers, groups known to havelower risk for developing cardiovascular disease, had signifi-cantly higher LCAT mass concentrations than men or smokers,respectively. The transgenic rabbits in this study expressedhuman LCAT at 10-fold the median value for man. Thus,enhanced LCAT activity may decrease atherogenesis andrepresent a "no fault" (53) approach to therapy in man.

In conclusion, HDLs are now recognized as key plasmalipoproteins in modulating the development of atherosclerosis.The present studies demonstrate that overexpression of LCATresults in an antiatherogenic lipoprotein profile with reducedapolipoprotein B-containing lipoproteins in addition to ele-vated plasma levels of HDL. These changes were associatedwith markedly reduced atherosclerosis. These combined re-

sults indicate that modulation of plasma LCAT activity rep-resents a new target for developing pharmacologic and genetherapy strategies to prevent atherosclerosis.

We thank Dr. Robert Wall at the U.S. Department of Agriculturefor his many helpful discussions in establishing our transgenic rabbitprogram and Ms. Donna James for her efforts in the preparation of themanuscript. This work was supported, in part, by National Institutes ofHealth Grants HL54246-01, JL30086, and HL45694.

1. Gordon, D. J. & Rifkind, B. M. (1989) N. Engl. J. Med. 321, 1311-1316.2. Gordon, D. J., Knoke, J., Probstfield, J. L., Superko, R. & Tyroler, H. A.

(1986) Circulation 74, 1217-1225.3. Glomset, J. A., Janssen, E. T., Kennedy, R. & Dobbins, J. (1966) J. Lipid

Res. 7, 638-648.4. Glomset, J. A. (1968) J. Lipid Res. 9, 155-167.5. Miller, G. J. & Miller, N. E. (1975) Lancet i, 16-19.6. Rhoads, G. G., Gulbrandsen, C. L. & Kagan, A. (1976) N. Engl. J. Med. 294,

293-298.7. Gordon, T., Castelli, W P., Hjortland, M. C., Kannel, W. B. & Dawber,

T. R. (1977) Am. J. Med. 62, 707-714.8. Miller, N. E., Nestel, P. J. & Clifton-Bligh, P. (1977) Adv. Exp. Med. Biol.

82, 294-296.

9. Berg, K. & Borresen, A. L. (1976) Lancet i, 499-501.10. Hoeg, J. M. & Remaley, A. T. (1994) in Genetic Factors in Coronary Heart

Disease, eds. Goldbourt, U., de Faire, U. & Berg, K. (Kluwer, Boston), pp.351-369.

11. Assmann, G., von Eckardstein, A. & Funke, H. (1991) Curr. Opin. Lipidol.2, 110-117.

12. Klein, H.-G., Santamarina-Fojo, S., Duverger, N., Clerc, M., Dumon,M.-F., Albers, J. J., Marcovina, S. & Brewer, H. B., Jr. (1993)J. Clin. Invest.92, 479-485.

13. Klein, H.-G., Duverger, N., Albers, J. J., Marcovina, S., Brewer, H. B., Jr.,& Santamarina-Fojo, S. (1995) J. Biol. Chem. 270, 9443-9447.

14. Brewer, H. B., Jr., Santamarina-Fojo, S. & Hoeg, J. M. (1995) in Endocri-nology, eds. DeGroot, L. J., Besser, M., Jameson, J. L., Loriaux, D.,Marshall, J. C., Odell, W. D., Potts, J. T., Jr., Rubenstein, A. H., Cahill,G. F., Jr., Martini, L. & Nelson, D. H. (Saunders, Philadelphia), pp.2731-2753.

15. Vaisman, B. L., Klein, H.-G., Rouis, M., Berard, A. M., Kindt, M. R.,Talley, G. D., Meyn, S. M., Hoyt, R. F., Jr., Marcovina, S. M., Albers, J. J.,Hoeg, J. M., Brewer, H. B., Jr., & Santamarina-Fojo, S. (1995) J. Biol.Chem. 270, 12269-12275.

16. Francone, 0. L., Gong, E. L., Ng, D. S., Fielding, C. J. & Rubin, E. M.(1995) J. Clin. Invest. 96, 1440-1448.

17. Hoeg, J. M., Vaisman, B. L., Demosky, S. J., Jr., Meyn, S. M., Talley, G. D.,Hoyt, R. F., Jr., Feldman, S., Berard, A. M., Sakai, N., Wood, D., Brous-seau, M. E., Marcovina, S., Brewer, H. B., Jr., & Santamarina-Fojo, S.(1996) J. Biol. Chem. 271, 4396-4402.

18. Anitschkow, N. & Chalatow, S. (1913) Zentralbl. Allg. Pathol. Pathol. Anat.24, 1-9.

19. Klein, H.-G., Lohse, P., Duverger, N., Albers, J. J., Rader, D. J., Zech,L. A., Santamarina-Fojo, S. & Brewer, H. B., Jr. (1993) J. Lipid Res. 34,49-58.

20. Albers, J. J., Tollefson, J. H., Chen, C. H. & Steinmetz, A. (1984) Arterio-sclerosis (Dallas) 4, 49-58.

21. Iverius, P.-H. & Brunzell, J. D. (1985) Am. J. Physiol. 249, E107-E1 14.22. Warnick, G. R., Benderson, J., Albers, J. J., Baillie, E. E., Sexton, B.,

Schaefer, E. J., Carlson, D., Hill, M., Brewer, H. B., Jr., Wiebe, D. A.,Hazlehurst, J. & Cooper, G. R. (1982) Clin. Chem. 28, 1379-1388.

23. Jiao, S., Cole, T. G., Kitchens, R. T., Pfleger, B. & Schonfeld, G. (1990)Metabolism 39, 155-160.

24. Kolodgie, F. D., Virmani, R., Cornhill, J. F., Herderick, E. E., Malcom,G. T. & Mergner, W. J. (1992) Atherosclerosis (Berlin) 97, 53-62.

25. Chobanian, A. V., Lichtenstein, A. H., Nilakhe, V., Haudenschild, C. C.,Drago, R. & Nickerson, C. (1989) Hypertension 14, 203-209.

26. Folch, R. J., Lees, M. & Sloan-Stanley, G. H. (1957) J. Biol. Chem. 226,497-509.

27. Lowry, 0. H., Rosebrough, N. J., Farr, A. L. & Randall, R. J. (1951) J. Biol.Chem. 193, 265-275.

28. Castelli, W. & Leaf, A. (1985) Cardiol. Clin. 3, 171-178.29. Castelli, W. P., Garrison, R. J., Wilson, P. W., Abbott, R. D., Kalousdian,

S. & Kannel, W. B. (1986) J. Am. Med. Assoc. 256, 2835-2838.30. Quinet, E. M., Agellon, L. B., Kroon, P. A., Marcel, Y. L., Lee, Y. C.,

Whitelock, M. E. & Tall, A. R. (1990) J. Clin. Invest. 85, 357-363.31. Borresen, A. L. (1976) J. Immunogenet. 3, 73-81.32. Borresen, A. L. (1976) J. Immunogenet. 3, 83-89.33. Borresen, A. L. (1976) J. Immunogenet. 3, 91-103.34. Ha, Y. C. & Barter, P. J. (1982) Comp. Biochem. Physiol. B 71, 265-269.35. Marzetta, C. A., Meyers, T. J. & Albers, J. J. (1993) Arterioscler. Thromb.

13, 834-841.36. Tall, A. R. (1993) J. Lipid Res. 34, 1255-1274.37. Badimon, J. J., Badimon, L. & Fuster, V. (1990) J. Clin. Invest. 85,

1234-1241.38. Rubin, E. M., Krauss, R. M., Spangler, E. A., Verstuyft, J. G. & Clift, S. M.

(1991) Nature (London) 353, 265-267.39. Plump, A. S., Scott, C. J. & Breslow, J. L. (1994) Proc. Natl. Acad. Sci. USA

91, 9607-9611.40. Rader, D. J., Schaefer, J. R., Lohse, P., Ikewaki, K., Thomas, F., Harris,

W. A., Zech, L. A., Dujovne, C. A. & Brewer, H. B., Jr. (1993) Metabolism42, 1429-1434.

41. Barbaras, R., Puchois, P., Fruchart, J. C. & Ailhaud, G. (1987) Biochem.Biophys. Res. Commun. 142, 63-69.

42. de la Llera-Moya, M., Atger, V., Paul, J. L., Fournier, N., Moatti, N., Giral,P., Friday, K. E. & Rothblat, G. (1994)Arterioscler. Thromb. 14,1056-1065.

43. Gospodarowicz, D. & Cheng, J. (1987) In Vitro Cell Dev. Biol. 23, 507-514.44. Parthasarathy, S., Barnett, J. & Fong, L. G. (1990) Biochim. Biophys. Acta

1044, 275-283.45. Barbaras, R., Puchois, P., Grimaldi, P., Barkia, A., Fruchart, J. C. &

Ailhaud, G. (1988) Adv. Exp. Med. Biol. 243, 271-277.46. Fruchart, J. C. & Ailhaud, G. (1992) Clin. Chem. 38, 793-797.47. Brewer, H. B., Jr., Rader, D. J., Fojo, S. S. & Hoeg, J. M. (1990) in Disorders

of HDL, ed. Carlson, L. A. (Smith-Gordon, London), pp. 51-58.48. Cheung, M. C. & Wolf, A. C. (1989) J. Lipid Res. 30, 499-509.49. Ross, R. (1993) Nature (London) 362, 801-809.50. Glomset, J. A., Assmann, G., Gjone, E. & Norum, K. R. (1995) in The

Metabolic and Molecular Bases of Inherited Disease, eds. Scriver, C. R.,Beaudet, A. L., Sly, W. S., Valle, D., Stanbury, J. B., Wyngaarden, J. B. &Fredrickson, D. S. (McGraw-Hill, New York), pp. 1933-1951.

51. Kuivenhoven, J. A., Stalenhoef, A. F. H., Hill, J. S., Demacker, P. N. M.,Errami, A., Kastelein, J. J. P. & Pritchard, P. H. (1996) Arterioscler.Thromb. Vasc. Biol. 16, 294-303.

52. Albers, J. J., Bergelin, R. O., Adolphson, J. L. & Wahl, P. W. (1982)Atherosclerosis (Berlin) 43, 369-379.

53. Roth, J. (1991) in Frontiers in Diabetes Research: Lessons From AnimalDiabetes II, ed. Shafrir, E. (Smith-Gordon, London), pp. 325-333.

Applied Biological Sciences: Hoeg et al.